China Maddison Wu 22 Feb 07

Prévia do material em texto

1 
CHINA’S ECONOMIC PERFORMANCE: HOW FAST HAS GDP GROWN; 
HOW BIG IS IT COMPARED WITH THE USA?* 
By Angus Maddison and Harry X. Wu 
 In the past two decades, China has been the world’s fastest growing economy. 
In 1982, it overtook Germany as the world’s third largest economy; in 1992, overtook 
Japan as the second biggest economy. In 2003 its GDP was about 73 per cent of that 
in the USA. It seems likely that it will overtake the USA, and become number one 
before 2020. 
 The above estimates were made by converting China’s GDP and that of the 
other countries in their national currencies, using a PPP (purchasing power parity) 
converter, rather than the exchange rate. As the Chinese currency is greatly 
undervalued, the difference between PPP and exchange rate conversion is unusually 
large. With exchange rate conversion, Chinese GDP appears to have been only 15 per 
cent of that in the USA in 2003. Very frequently, there is significant error in assessing 
China’s comparative performance when such exchange rate comparisons are used. 
This happens in journalism, in political discourse and also amongst some economists 
who regard Japan as the second largest economy although it now has a GDP less than 
half the Chinese. In 2003, German GDP was about a quarter of that in China. 
Estimation of the PPP converter for China is described in Maddison (1998) 
Appendix C, pp. 153-154. It was derived from the China/US expenditure comparison 
for 1986 in Ren (1997), China’s Economic Performance in International Perspective, 
OECD, Paris, p. 37. His estimates were updated to 1990 which is the benchmark year 
for all other countries in Maddison (1995, 2001, 2003 and www.ggdc.net/Maddison). 
Here the PPP estimate of China’s GDP in 1990 has been raised from $2,109,400 to 
$2,123,852, as the estimated 1986-1990 change in GDP volume between these years 
has been adjusted very slightly upwards. 
(I) Official Estimates of Chinese GDP Growth 
 Here we are mainly concerned with a different problem: adjustment of the 
official Chinese estimates of GDP growth. These were initially based on the Soviet 
material product system (MPS) which greatly exaggerated growth performance. They 
are now based on the western SNA system; they still exaggerate growth, but to a 
much smaller degree. 
 Until the late 1980s, China used the Soviet material product system of 
accounts which systematically overstated growth, and excluded a large part of service 
activity. There were also serious deficiencies in the basic reporting system. In the 
turbulence of the cultural revolution, the statistical office was abolished in 1968 and 
its staff disbursed. It was re-established in 1972, but most of the old personnel had 
disappeared and many old records had been destroyed. No new graduates with the 
requisite training had been produced in the years when the universities were closed. 
 The statistical office published two measures of aggregate economic 
performance. The “total product of society” showed average annual growth rate in 
“comparable prices” of 7.9 per cent for 1952-78 (1988 Yearbook p. 28). It referred to 
aggregate gross output of five sectors and involved a good deal of double counting 
because each of the component sectors had significant inputs from the others. “Net 
material product” which the Chinese called “national income” showed 6 per cent 
growth for the same period at “comparable prices” (p. 42 of the 1988 Yearbook); this 
was better as it deducted most inputs except “non-material services”. 
 
*We are grateful for comments by Ximing Yue on an earlier version presented in December 2006 to a 
University of Queensland seminar on World Economic Performance: Past, Present and Future. 
 2 
 In the early 1990s, these Soviet style measures were dropped in favour of 
western concepts showing slower growth. The statistical office also stopped reporting 
estimates for the Maoist period 1952-78; since 1995, the Yearbooks have only 
published estimates for 1978 onwards. However, there was a joint retrospective 
exercise by the statistical office and Hitotsubashi University in 1997 which provided 
western type estimates for 1952-1995. This exercise showed a growth rate of GDP of 
4.7 per cent a year for 1952-78. The Maddison-Wu revision shows a growth rate of 
4.4 for this period. For 1978-2003 Maddison–Wu show growth of 7.85 per cent a year 
compared with the official 9.6 per cent. Table 1 summarises the difference between 
the revised Maddison-Wu estimates and those of the NBS. 
TABLE 1 OFFICIAL & MADDISON-WU ESTIMATES OF GDP AND GDP PER CAPITA 
(1990 international Geary-Khamis $) 
 
Official 
GDP 
 
Population 
 
 
Official 
per Capita 
GDP 
 
GDP Revised 
by 
Maddison-
Wu 
Revised 
Maddison-Wu 
per capita GDP 
US GDP 
 
 (million $) (000s) ($) (million $) ($) (million $) 
1952 231,550 568,910 407 305,854 538 1,625,245 
1978 769,694 956,165 805 935,083 978 4,089,548 
1990 2,123,852 1,135,185 1,871 2,123,852 1,871 5,803,200 
2003 7,598,267 1,288,400 5,897 6,187,983 4,803 8,414,751 
Source: Last columns of Tables 8 and 9 for China. www.ggdc.net/ Maddison for USA. 
 The estimates of the statistical office for 1978 onwards are a great 
improvement on the earlier MPS system. They are made independently from the 
production and expenditure side, but the former are considered more reliable (see Xu 
and Ye, 2000, p. 12). They are based on the international standard SNA guidelines, 
but there is still scope for improvement. The national accounts department of the 
National Bureau of Statistics (NBS) is of high quality, but relatively small. In 2000, 
there were only 26 staff members. Reports at the basic level reach NBS through 
several levels of aggregation in the administrative hierarchy. This transmission train 
provides opportunities for officials at different levels to adjust their reports to reflect 
favourably on their management. NBS makes crosschecks, but they are necessarily 
limited in scope. Xu and Ye give a detailed account of the procedures used in 
constructing official accounts. Except for two sectors (agriculture and transport), 
growth measures are not based on quantitative indicators of volume movement. Most 
sector accounts are constructed at current prices and deflated by price indices. 
Because of the shakiness of the price reports it receives, the NBS, until 2002, 
distinguished between current and “comparable” prices instead of current and 
“constant” prices as other countries do. State enterprises use a price manual specifying 
2,000 items to help them differentiate current and constant prices. Many reporting 
units in the private sector cannot or do not make this distinction. This is an important 
reason for overstatement in the official statistics. 
(II) Maddison-Wu Re-estimation of Chinese Growth, 1952-2003 
 Maddison (1998) re-estimated Chinese performance to produce a measure of 
growth and levels closer to western SNA practice. This required recalculation of the 
annual change in real GDP at 1987 prices from 1952 to 1995. Maddison (2001 and 
2003) contained rough updates of the GDP growth estimate beyond 1995, using a 
“correction coefficient” derived by applying the ratio (.7586) of the Maddison 
estimates to the official growth rate for 1978-95 to correct the official figures beyond 
1995. It is now clear that this type of updating is unsatisfactory, as the difference 
between the official and the Maddison estimates is not stable from year to year. In this 
 3 
paper, Maddison and Wu have made comprehensive revisions and updates for 1952-
2003 which supersede earlier Maddison estimates. 
 (1) Maddison (1998) made his own estimates of gross value added in farming, 
using price and quantity data of the Food and Agriculture Organisation for 125 crop 
and livestock items, adjusted for changes in farm and non-farminputs. For fishery, 
forestry and agricultural sidelines he used the official estimates. He found 
approximately the same rate of growth as the official estimates for the agricultural 
sector as a whole in 1952-90, but his level of value added was nearly one fifth higher 
in 1990. In view of the close congruence of the official and Maddison estimates up to 
1990, the official estimates for this sector have been used to update the Maddison 
estimates from 1991 to 2003. 
 (2) for industry, Wu’s (1997) estimates of gross value added in industry were 
used in Maddison (1998). This was a volume index, with detailed time series on 
physical output and prices from the China Industrial Economic Statistical Yearbook. 
Value added was derived from the official input-output table. Wu (2002) presented a 
bigger sample covering 117 products, and explained in detail why the official figures 
exaggerated growth. He provided detailed time series showing annual movement for 
15 branches of manufacturing as well as mining and utilities. Here he has revised and 
updated these estimates to 2003. He shows a growth rate of 10.1 per cent a year for 
industry as a whole for 1952-78, compared to the official 11.5 per cent; and 9.75 per 
cent a year for 1978-2003 compared to the official 11.5 per cent. (see tables 3, 4 and 6 
below). Figure 2 compares the time-profile of the official and the Wu estimates for 
industry. The profile of the two estimates is very different in the 1990s, with Wu 
showing a distinct slowdown of growth in 1996-98, and faster growth than the official 
estimates thereafter. It seems clear that the official estimates for industry involved a 
smoothing of the growth path in the 1990s. 
 (3) In the old Soviet-style national accounts “non-material services” were 
excluded from “material product”. These are banking, insurance, housing services, 
administration of real estate, social services, health, education, entertainment, 
personal services, R & D activities, the armed forces, police, government and party 
organisations. They are now incorporated in the Chinese accounts, but the estimates 
are not shown explicitly. Official estimates show an estimate for the “tertiary” sector 
as a whole, and a breakdown for two component sub-sectors (transport and 
commerce). The estimate for non-material services is a residual which the reader has 
to derive for him or herself. As GDP originating in this sub-sector is bigger than in 
transport or commerce, it suggests that the residual treatment reflects official qualms 
about its measure for this sector. 
 The international standardised System of National Accounts (1993, p. 134), 
recommends valuation of non-market output by the cost of labour input minus 
intermediate consumption. NBS credits these activities with improbably high rates of 
growth of labour productivity (1.5 per cent per annum per person employed for 1952-
78 and 5.1 per cent for 1978-2003). Maddison (1998) made a rough alternative 
estimate, assuming no increase in productivity in these activities. He used 
employment as a proxy indicator of growth in real value added, and the same 
assumption is made in the present estimates. He used the employment estimates in the 
NBS Statistical Yearbooks, adjusted them from an end-year to a mid-year basis, and 
added a proxy estimate of 3 million a year for the military which were not included in 
the official figures until 1992 (see Maddison 1998, p, 171). Here these Maddison 
estimates have been updated to 2003 (see Table 8). 
 4 
(4) for construction, Maddison (1998) used the official figures for 1952 onwards, and 
this has been done here. 
(5) for the sectors “transport and communications” and “commerce and restaurants”, 
official volume figures were used for 1957 onwards. For 1952-57, Liu and Yeh (1965) 
estimates were used because they included estimates for traditional modes of transport 
and for the activity of pedlars which seem to have been omitted from the official 
estimates. In 2006, the NBS made a significant upward adjustment to its earlier 
estimates of growth in these two sectors for 1993 onwards (see section III below). 
 The need for adjustment of the Chinese official estimates was acknowledged 
by official statisticians (Xu Xianchun, Deputy Commissioner of NBS (National 
Bureau of Statistics), and Ye Yanfei, former head of the social division): 
“There can be no doubt that China’s official national accounts are regarded with 
suspicion by many users outside China. Professor Maddison’s 1998 study raised 
serious questions about both the levels and growth rates of China’s GDP. His 
criticism carries particular weight because it is based on a careful reworking of the 
GDP estimates and because he consulted widely with Chinese scholars who have 
first-hand knowledge of economic developments in China.” “ the sheer size of China, 
together with the limited resources currently devoted to national accounts and the 
continuation of MPS-oriented statistical procedures, inevitably means that the official 
GDP estimates are subject to margins of error that are somewhat bigger than for 
other developing countries and substantially larger compared with most other OECD 
countries” “A reasonable assessment might be that the official growth estimates 
represent an upper bound and the Maddison estimates a lower bound, with the true 
growth rates lying somewhere between the two”. 
(III) Recent Upward Revision of Official Estimates for the Service Sector 
 The Chinese statistical office produces two types of basic survey which 
provide crosschecks on the accuracy of the national accounts. The first input-output 
estimates for the whole economy appeared in 1987. They were used to estimate sector 
value added in Maddison (1998). Since then, condensed input-output tables were 
published for 1990, 1995 and 2000, and a full-scale set of tables for 1992, 1997 and 
2002. 
 The other major source of information is the census of economic activity. A 
first industrial census was taken in the 1950s, a second for 1985, and a third for 1995. 
There was a census for the service sector in 1993 and a rather slim census for 
agriculture in 1996. The latest census for 2004 covered the whole economy excluding 
agriculture. Its results were published in four volumes in January 2006. The 
respondents had to supply information on the value of their sales between January 1 
and December 31 of 2004 at current prices; estimates of their capital stock (at book 
value), employment, wages, taxes, debt, but nothing on value added. In the light of 
this new information, the statistical office made a substantial upward adjustment in its 
measure of GDP growth in the two service sectors transport and communication and 
commerce for 1993-2004. We have accepted these revised official estimates. 
However, we have some reservations, as the updating was done with current price 
information, and it is not clear whether all of the increase should have been assumed 
to occur after 1992. 
 The revised estimates were presented in the 2006 Yearbook. NBS calculated a 
new trend in the value of output in current prices for the years 1992 and 2004, and 
compared it with the trend based on its former estimates for this period. The ratio of 
the two trends was used to inflate the former estimates of the year-to-year change in 
the component series of real GDP. Table 2 shows the outcome of this procedure in 
 5 
real terms-it involved an increase annual average real GDP growth from 9.4 to 9.9 per 
cent for 1992-2003, the new Maddison-Wu estimate shows an 8.7 growth rate. 
 The bottom panel of Table 2 shows the Maddison-Wu estimates. The columns 
for industry, non-material services and GDP differ markedly from the NBS estimates. 
TABLE 2 REAL SECTORAL GDP GROWTH RATES: OLD AND REVISED OFFICIAL AND 
MADDISON-WU ESTIMATES, 1992-2003 
(Annual average percentage growth rates)GDP 
 
Agriculture 
 
Industry 
 
Construction 
 
Transportation, 
Communication 
Commerce 
 
Non-Material 
Services 
 Original NBS 
1992 14.2 4.7 21.2 21.0 10.5 13.1 12.8 
1993 13.5 4.7 20.1 18.0 12.4 6.6 11.8 
1994 12.6 4.0 18.9 13.7 9.5 7.7 10.4 
1995 10.5 5.0 14.0 12.4 12.0 5.9 8.1 
1996 9.6 5.1 12.5 8.5 11.4 5.4 7.6 
1997 8.8 3.5 11.3 2.6 10.8 8.5 8.7 
1998 7.8 3.5 8.9 9.0 10.6 7.7 7.6 
1999 7.1 2.8 8.5 4.3 11.3 7.2 6.5 
2000 8.0 2.4 9.8 5.7 11.5 8.2 6.7 
2001 7.5 2.8 8.7 6.8 9.5 7.5 8.3 
2002 8.3 2.9 10.0 8.8 7.9 8.1 9.3 
2003 9.5 2.5 12.8 12.1 6.3 9.1 8.0 
1992-2003 9.4 3.6 12.3 9.2 10.3 7.4 8.4 
 Revised NBS 
1992 14.2 4.7 21.2 21.0 10.5 13.1 12.8 
1993 14.0 4.7 20.1 18.0 14.5 8.4 12.8 
1994 13.1 4.0 18.9 13.7 11.6 9.5 11.4 
1995 10.9 5.0 14.0 12.4 14.1 7.7 9.1 
1996 10.0 5.1 12.5 8.5 13.6 7.2 8.7 
1997 9.3 3.5 11.3 2.6 12.9 10.4 10.0 
1998 7.8 3.5 8.9 9.0 10.6 7.8 7.6 
1999 7.6 2.8 8.5 4.3 13.4 9.1 7.7 
2000 8.4 2.4 9.8 5.7 13.6 10.1 7.9 
2001 8.3 2.8 8.7 6.8 11.6 9.3 9.9 
2002 9.1 2.9 10.0 8.8 9.9 10.0 10.8 
2003 10.0 2.5 12.8 12.1 8.3 11.0 9.5 
1992-2003 9.9 3.6 12.3 9.2 12.2 9.1 9.6 
 Maddison-Wu 
1992 9.7 4.7 13.1 21.0 10.5 13.1 6.1 
1993 9.7 4.7 13.8 18.0 14.5 8.4 5.1 
1994 10.0 4.0 12.9 13.7 11.6 9.5 10.8 
1995 15.1 5.0 26.1 12.4 14.1 7.7 7.9 
1996 2.1 5.1 -3.2 8.5 13.6 7.2 4.3 
1997 5.3 3.5 5.1 2.6 12.9 10.4 3.9 
1998 0.3 3.5 -7.1 9.0 10.6 7.8 4.3 
1999 6.6 2.8 9.6 4.3 13.4 9.1 1.4 
2000 9.0 2.4 14.5 5.7 13.6 10.1 3.1 
2001 10.7 2.8 17.3 6.8 11.6 9.3 4.8 
2002 12.4 2.9 20.7 8.8 9.9 10.0 3.7 
2003 15.1 2.5 24.9 12.1 8.3 11.0 4.5 
1992-2003 8.7 3.6 11.8 9.2 12.2 9.1 4.9 
Source: Original NBS estimates are from China Statistical Yearbook 2005, p.53, Table 3.3, revised NBS estimates 
from China Statistical Yearbook 2006, p.59-60, Tables 3.3 and 3.4, Maddison-Wu estimates from Table 8. 
Note: Figures in bold italics in the middle panel indicate differences from the original NBS estimates. Figures in 
bold italics in the bottom panel indicate differences from the revised NBS revised estimates. 
 
(IV)Wu’s Alternative Estimates of Industrial Performance 
 NBS derives its aggregate measure of industrial performance by cumulating 
returns from enterprises with very few independent checks. These returns show output 
in current and in “comparable” prices. In order to measure “comparable” prices, 
enterprises are given price manuals which specify the prices they are to use for 
benchmark years. However the specification manuals do not cover all items produced 
 6 
or specify in sufficient detail. State enterprises have an incentive to exaggerate 
performance by understating inflation. Although there are penalties for falsification, 
there are substantial possibilities for exaggerating the volume of output when new 
products are incorporated into the reporting system at so-called “comparable” prices.. 
Many of the new small-scale non-state enterprises cannot or do not bother to 
distinguish between current and “comparable” prices, so the tendency to understate 
inflation has increased 
 Wu (1997, 2000, and 2002) made alternative estimates of industrial value 
added in constant 1987 prices for 1952 onwards which are much better than the 
official figures for several reasons. They are based on physical indicators for a 
relatively large number of products from the official Industrial Economic Statistics 
Yearbooks, with 1987 value added weights from NBS, Input Output Table of China 
1987 (199l, pp.147-62). The exercise is fully transparent and follows methods used in 
Western countries. The procedure is rather like that for measuring farm output, except 
that it was not possible to adjust for possible changes in input ratios over time. The 
coverage corresponds to that in Western definitions (ie. it excludes forestry products 
and repair and maintenance which are included in the official statistics for industry). 
There is a breakdown for 17 sectors: mining, utilities, and 15 manufacturing sectors, 
following the standard industrial classification. Table 3 provides a comparative 
picture of the pace of advance in different sectors and the locus of structural change. 
TABLE 3 WU'S RATES OF GROWTH AND SHARES OF VALUE ADDED BY 
INDUSTRIAL BRANCH, 1952–2003 
 Growth Rates 
(annual average compound 
growth rates) 
Branch Share of Gross Value Added 
(per cent) 
 1952–78 1978–2003 1952 1978 2003 
 
Food Products 6.4 9.5 6.3 2.6 2.4 
Beverages 9.1 7.4 1.6 1.3 0.7 
Tobacco Products 5.9 4.5 10.0 3.6 1.1 
Textile Products 5.9 6.3 27.5 9.8 4.5 
Wearing Apparel 3.8 16.9 5.7 1.2 5.9 
Leather Goods and Footwear 8.4 10.6 1.2 0.8 1.0 
Wood Prods, Furniture & Fixtures 6.3 5.3 6.2 2.4 0.9 
Paper, Printing & Publishing 10.0 10.1 2.5 2.5 2.7 
Chemical & Allied Products 13.7 8.6 6.7 15.5 11.8 
Rubber & Plastic Products 11.5 11.7 1.8 2.4 3.7 
Non-Metallic Mineral Products 9.5 9.9 7.0 6.0 6.2 
Basic & Fabricated Metal Products 15.7 5.7 2.5 9.0 3.5 
Machinery & Transport Equipment 16.4 9.7 3.2 13.4 13.2 
Electrical Machinery & Equipment 16.5 18.0 1.0 4.4 27.1 
Other 9.2 13.6 4.2 3.3 7.9 
 
Total Manufacturing 9.7 10.5 87.4 78.3 92.6 
Mining 11.9 3.0 10.6 16.0 3.3 
Utilities 14.7 8.4 2.0 5.8 4.2 
Total Industry 10.1 9.8 100.0 100.0 100.0 
Source: H. X. Wu (2002), “How Fast has Chinese Industry Grown? Measuring the Real Output of 
Chinese Industry, 1949-97” Review of Income and Wealth, (48), no. 2, pp. 179-204, updated. 
 
 
 7 
TABLE 4 
NEW WU ESTIMATES OF INDUSTRIAL VALUE ADDED 
1952-2003 
 
Industrial GDP in 1987 Yuan 
 (Million) 
 Manufacturing Mining Utilities Total 
1952 15,548 1,889 359 17,796 
1953 20,246 1,916 452 22,615 
1954 23,427 2,405 541 26,373 
1955 23,912 3,099 605 27,616 
1956 29,920 3,190 816 33,926 
1957 33,116 4,155 949 38,220 
1958 49,262 7,955 1,352 58,569 
1959 66,796 10,717 2,080 79,592 
1960 71,770 12,465 2,920 87,155 
1961 36,504 8,409 2,360 47,272 
1962 31,978 6,760 2,252 40,990 
1963 38,100 6,930 2,409 47,439 
1964 46,841 6,725 2,753 56,319 
1965 57,992 8,587 3,324 69,902 
1966 69,882 9,612 4,056 83,550 
1967 57,385 8,433 3,805 69,624 
1968 54,205 8,979 3,520 66,705 
1969 71,884 10,730 4,622 87,236 
1970 94,673 14,742 5,698 115,113 
1971 104,960 17,672 6,804 129,437 
1972 112,177 19,469 7,493 139,139 
1973 123,493 20,494 8,201 152,188 
1974 118,354 22,704 8,299 149,357 
1975 134,009 26,670 9,627 170,306 
1976 129,346 28,053 9,985 167,385 
1977 147,732 31,252 10,984 189,968 
1978 171,695 35,003 12,616 219,314 
1979 188,271 34,803 13,865 236,940 
1980 199,233 34,825 14,779 248,837 
1981 204,491 33,860 15,207 253,557 
1982 219,478 34,653 16,111 270,242 
1983 237,326 36,455 17,277 291,058 
1984 265,054 39,649 18,535 323,239 
1985 302,008 41,842 20,192 364,042 
1986 328,943 43,931 22,100 394,974 
1987 390,072 46,420 24,450 460,943 
1988 425,387 48,852 26,805 501,043 
1989 437,587 51,552 28,752 517,891 
1990 446,245 51,622 30,542 528,408 
1991 498,600 53,056 33,310 584,966 
1992 569,410 54,979 37,066 661,456 
1993 655,098 56,253 41,274 752,626 
1994 744,676 59,452 45,630 849,759 
1995 956,156 65,472 49,510 1,071,138 
1996 916,307 66,963 53,162 1,036,432 
1997 965,662 68,257 55,832 1,089,751 
1998 893,110 62,284 57,376 1,012,770 
1999 990,267 58,789 60,931 1,109,987 
2000 1,146,735 57,324 66,649 1,270,707 
2001 1,356,913 61,043 72,804 1,490,760 
2002 1,651,151 66,513 81,319 1,798,983 
2003 2,079,632 73,222 93,935 2,246,790 
 
 8 
 FIGURE 1 
ESTIMATES OF INDUSTRIAL VALUE ADDED: NBS AND WU ESTIMATES 
1952-2003 
1.0
10.0
100.0
1000.0
19
52
19
55
19
58
19
61
19
64
19
67
19
70
19
73
19
76
19
79
19
82
19
85
19
88
19
91
19
94
19
97
20
00
20
03
Wu (1990=100)
Off icial (1990=100)
-60.0
-40.0
-20.0
0.0
20.0
40.0
60.019
52
19
55
19
58
19
61
19
64
19
67
19
70
19
73
19
76
19
79
19
82
19
85
19
88
19
91
19
94
19
97
20
00
20
03
Wu (% p.a.)
Official (% p.a.)
. 
The time-profile of the Wu estimates of industrial value added in Figure 1 shows 
slower growth than the official estimates for 1996-1999, and significantly faster 
growth thereafter. This suggests that the official estimates of growth performance in 
recent years have been smoothed. 
 (V) Alternative Estimates for Non-Material Services 
 In the old Soviet-style national accounts so-called “non-material services” 
were excluded from “material product”. They are now incorporated in the Chinese 
accounts. All are measurement resistant. NBS credits these activities with labour 
productivity growth averaging 3.2 percent a year from 1952 to 2003), but the 
international standardised System of National Accounts (1993, p. 134), recommends 
valuation of non-market output by the cost of labour input minus intermediate 
consumption. Table 5 shows the situation in OECD countries, where the average 
practice is to assume virtually zero productivity growth in this sector: 
TABLE 5 GDP PER PERSON EMPLOYED IN OECD COUNTRIES, 1973-90 
 (annual average compound growth rates) 
 Agriculture Industry Non-material Other 
 
 services services 
Denmark 6.42 2.24 0.26 1.76 
France 5.22 3.01 0.98 1.84 
Germany 5.48 1.83 1.00 2.62 
Italy 3.35 3.14 0.00 1.12 
Netherlands 4.25 1.63 -1.00 1.60 
Spain 6.26 4.74 1.35 2.15 
Sweden 3.84 2.12 -1.60 1.71 
UK 3.77 2.79 0.57 1.25 
USA 2.95 1.20 -1.00 0.77 
Average 4.62 2.52 0.06 1.65 
Source: van Ark (1996), pp. 109-115. 
 
 
 9 
TABLE 6 
MADDISON ESTIMATES OF “NON-MATERIAL” SERVICE EMPLOYMENT AND OUTPUT 
(1998 UPDATED) 
(Employment in 000s at mid-year, and GDP in 1987 yuan) 
 
Civilian 
Employment 
Assumed 
Military 
Manpower 
Total 
Employment 
in sector 
Sectoral 
GDP 
Maddison 
Million 
1987 yuan 
Sectoral 
GDP 
Official 
Million 
1987 yuan 
 
 
1952 7,023 3,000 10,023 45,486 13,879 
1953 7,365 3,000 10,365 47,038 16,597 
1954 7,580 3,000 10,580 48,014 15,336 
1955 7,754 3,000 10,754 48,803 16,877 
1956 8,688 3,000 11,688 53,042 19,923 
1957 10,194 3,000 13,194 59,877 21,883 
1958 17,905 3,000 20,905 62,512 26,564 
1959 27,615 3,000 30,615 65,264 30,870 
1960 31,515 3,000 34,515 68,136 34,693 
1961 24,900 3,000 27,900 71,135 27,459 
1962 15,450 3,000 18,450 74,266 25,178 
1963 14,085 3,000 17,085 77,535 26,294 
1964 14,840 3,000 17,840 80,961 31,658 
1965 15,780 3,000 18,780 85,227 37,820 
1966 16,305 3,000 19,305 87,610 31,731 
1967 16,535 3,000 19,535 88,654 32,983 
1968 16,935 3,000 19,935 90,469 35,721 
1969 17,100 3,000 20,100 91,218 38,233 
1970 16,980 3,000 19,980 90,673 39,062 
1971 17,205 3,000 20,205 91,694 42,481 
1972 17,735 3,000 20,735 94,099 42,926 
1973 18,065 3,000 21,065 95,597 44,204 
1974 18,495 3,000 21,495 97,548 46,220 
1975 18,935 3,000 21,935 99,545 48,770 
1976 19,705 3,000 22,705 103,040 50,583 
1977 22,045 3,000 25,045 113,659 53,572 
1978 25,965 3,000 28,965 131,448 58,972 
1979 29,005 3,000 32,005 145,245 62,290 
1980 30,775 3,000 33,775 153,277 70,293 
1981 32,990 3,000 35,990 163,330 74,330 
1982 34,335 3,000 37,335 169,433 87,767 
1983 35,945 3,000 38,945 176,740 100,188 
1984 40,845 3,000 43,845 195,369 120,073 
1985 45,015 3,000 48,015 217,901 137,697 
1986 47,010 3,000 50,010 226,955 155,122 
1987 49,955 3,000 52,955 240,320 180,240 
1988 53,170 3,000 56,170 254,910 203,045 
1989 55,530 3,000 58,530 265,620 228,880 
1990 57,685 3,000 60,685 275,400 236,943 
1991 60,300 3,000 63,300 287,268 246,481 
1992 64,175 3,000 67,175 304,853 293,388 
1993 67,610 3,000 70,610 320,442 330,870 
1994 75,240 3,000 78,240 355,068 368,507 
1995 81,435 3,000 84,435 383,182 402,118 
1996 85,100 3,000 88,100 399,815 437,784 
1997 88,535 3,000 91,535 415,403 480,552 
1998 92,558 2,912 95,470 433,261 517,330 
1999 94,065 2,740 96,805 439,320 557,416 
2000 97,237 2,578 99,815 452,984 601,564 
2001 102,105 2,500 104,605 474,718 661,303 
2002 105,960 2,500 108,460 492,212 732,572 
2003 110,870 2,500 113,370 514,495 801,926 
NB: this table corrects errors for 1981 and 1985 employment in Maddison (1998), p. 171. Until 1992, the official 
figures excluded the military. It was assumed that the military are included in the official figures from 1993. 
Official figures are not available for total employment in 2003; 113,370 is an estimate of Ximing Yue. 
 
 10 
 Table 6 replicates and updates the procedure used in Maddison (1998) to 
provide an alternative measure of GDP in the “non-material” service sector, assuming 
no increase in productivity in these activities. Employment is used as a proxy 
indicator of growth in real value added. The employment estimates for this sector in 
the NBS Statistical Yearbooks were used, with two modifications: (i) adjusting the 
official employment estimates from an end-year to a mid-year basis (see first column), 
and (ii) adding a proxy estimate of 3 million a year for military personnel who were 
not included in the official estimates until 1993 (see China Labour Statistics Yearbook, 
1994). Maddison augmented the 1987 weight for this sector, as the official coverage 
appeared to be inadequate. It substantially undervalued housing and military outlays, 
and probably did not cover welfare benefits in kind which were provided free to 
employees of state enterprises. 
 The fourth column of Table 6 shows the Maddison estimates of GDP growth 
in this sector, in 1987 yuan. The annual movement is the same as for employment 
(shown in the third column). Column five shows the official estimates of GDP growth 
in this sector. They are also in 1987 prices. 
 Table 7 shows official employment statistics for the whole economy for 1952-
2002, with a fourfold sector breakdown which is the best we can do if we want a 
consistent picture for the whole period. The estimates for 1952-78 were derived from 
NBS, Statistical Yearbook of China (1993, Chinese version), pp. 100-101. It provided 
a four sector breakdown: a) for the primary sector; b) for the secondary sector; c) for 
the “material” service sector, and d) for the “non-material” service sector. For 1978-
2002, much more detail was available for sixteen branches. However the total 
coverage was consistent with the earlier estimates. In Table 7 they are aggregated into 
four sectors for the whole period. 
 The 16 branch breakdown included a consolidated estimate for agriculture, 
forestry, animal husbandry and fisheries; 4 separate components for the secondary 
sector-mining, manufacturing, construction and utilities (gas, water and electricity); 3 
components for “material” services-transport and communications, wholesale and 
retail trade, and geological prospecting; and 8 components for “non-material” activity. 
For 1978-85 estimates were derived from the 1994 Yearbook, p. 68; 1985-89 from the 
2000 Yearbook, pp. 120-121, 1989-2002 from the 2006 Yearbook, p. 130. 
 
 An Important Incongruity in the Official Estimates of Employment 
 Until 1997, The NBS had, in addition to the sixteen branch breakdown (pp. 
92-93), more aggregative employment estimates for 3 sectors, primary, secondary and 
tertiary (pp. 87). The figure for total employment was the same in the two tables. 
 In Yearbooks from 1997 onwards, there is a discrepancy between the two 
tables. Total employment in the 3 sector table is much bigger than for the 16 sectors. 
In the 2006 Yearbook (pp. 128 & 130), the 3 sector total for 1990 (end-year) was 
647.5 million and the actual total for the 16 sectors was 567.4 million. Hence a 
discrepancy of 80.1 million. For 2002, the discrepancy had risen to 99.6 million. 
Instead of explainingthe discrepancy, the Yearbooks disguised it by showing the 
same “total” for the 16 sector breakdown as for the three sector aggregate. 
 The 16 sector series continues to be published, but the figures stop at the year 
2002 in the last four Yearbooks. It would seem that the 3 sector breakdown is derived 
from the sample population census (see Yue, 2005) and the sixteen sector breakdown 
from labour force statistics, but users of the employment figures are entitled to a 
detailed explanation or reconciliation of the two types of estimate. They are also 
 11 
entitled to know why the 16 sector breakdown has been discontinued. In the present 
situation, meaningful measurement of labour productivity is no longer possible. 
 
TABLE 7 EMPLOYMENT BY SECTOR, OLD CLASSIFICATION, CHINA 1952–2003 
 (000s at mid–year) 
 Farming, Forestry, 
Fishery & Sidelines 
Industry and 
Construction 
“Material” 
Services 
“Non–Material” 
Services 
Total 
 1952 171 070 14 479 11 684 10 023 207 256 
1953 175 300 16 175 11 630 10 365 213 470 
1954 179 455 17 895 11 055 10 580 218 985 
1955 183 660 18 815 10 571 10 754 223 800 
1956 185 600 21 675 10 767 11 688 229 730 
1957 189 175 22 795 11 776 13 194 236 940 
1958 173 900 45 745 14 305 20 905 254 855 
1959 158 685 61 970 15 595 30 615 266 865 
1960 166 265 47 095 15 390 34 515 263 265 
1961 183 625 34 385 14 445 27 900 260 355 
1962 204 940 24 255 12 860 18 450 260 505 
1963 216 035 20 215 12 415 17 085 265 750 
1964 223 630 20 820 12 590 17 840 274 880 
1965 230 750 22 650 12 850 18 780 285 030 
1966 238 225 24 715 13 130 19 305 295 375 
1967 247 070 25 965 13 525 19 535 306 095 
1968 255 895 26 665 14 150 19 935 316 645 
1969 265 650 28 495 14 455 20 100 328 700 
1970 274 390 32 355 14 560 19 980 341 285 
1971 280 755 37 100 15 205 20 205 353 265 
1972 283 065 40 830 15 745 20 735 360 375 
1973 285 340 43 305 15 820 21 065 365 530 
1974 290 000 45 410 16 200 21 495 373 105 
1975 292 975 48 605 17 170 21 935 380 685 
1976 294 065 53 020 18 220 22 705 388 010 
1977 293 460 56 325 19 255 25 045 394 085 
1978 288 060 63 405 20 220 28 965 400 650 
1979 284 760 70 795 21 330 32 005 408 890 
1980 288 780 74 605 22 770 33 775 419 930 
1981 294 495 78 550 24 395 35,990 433 430 
1982 303 180 81 745 25 840 37 335 448 100 
1983 310 050 85 125 27 530 38 945 461 650 
1984 310 095 91 345 30 870 43,845 476 155 
1985 309 990 99 870 35 475 48 015 493 350 
1986 311 920 108 000 38 840 50 010 508 770 
1987 314 585 114 710 41 075 52 955 523 325 
1988 319 560 119 390 43 485 56 170 538 605 
1989 327 370 120 640 44 795 58,530 551,335 
1990 336 710 120 490 45,120 60 685 563,005 
1991 345 365 122 755 46,735 63 300 578,155 
1992 348 755 126 540 49,495 67,175 591,965 
1993 343 805 131 980 51,880 70,610 598,275 
1994 336 760 137 395 55,825 78,240 608,220 
1995 332,020 141 385 61,215 84,435 619,055 
1996 329,640 143,305 65,110 88,100 626,155 
1997 330,025 142,790 68,195 91,535 632,545 
1998 331,635 134,310 68,735 95,470 630,150 
1999 333,625 125,615 68,225 96,805 624,270 
2000 334,238 124,745 68,544 99,816 627,343 
2001 331,641 125,383 68,519 104,607 630,150 
2002 327,304 128,249 70,148 108,461 634,162 
 12 
(VI) Summary of Maddison-Wu Results 
 Figure 2 shows the difference between the official measure of GDP growth, 
and the Maddison-Wu estimates. Both are shown as volume indices benchmarked on 
1990. Tables 8 and 9 provide a full annual six-sector picture for 1952-2003. Here one 
can scrutinise the impact of the Maddison-Wu amendments in detail. Thus for the 
reform period 1978-2003, the official GDP growth rate is 9.59 per cent per annum. 
Replacing column 2 of table 9 for industry by column 2 of table 8 lowers the growth 
rate to 8.8 per cent. Replacing column 6 for non-material services reduces the growth 
rate further to 7.98. Replacing column 1 for agriculture reduces the growth rate to 
7.85. Thus the adjustment for non-material services has a slightly bigger impact than 
that for industry. 
FIGURE 2 GDP GROWTH: OFFICIAL AND MADDISON-WU ESTIMATES 1952-2003 
1.0
10.0
100.0
1000.0
19
52
19
55
19
58
19
61
19
64
19
67
19
70
19
73
19
76
19
79
19
82
19
85
19
88
19
91
19
94
19
97
20
00
20
03
Maddison-Wu (1990=100)
Off icial (1990=100)
-40.0
-30.0
-20.0
-10.0
0.0
10.0
20.0
30.0
40.0
19
52
19
55
19
58
19
61
19
64
19
67
19
70
19
73
19
76
19
79
19
82
19
85
19
88
19
91
19
94
19
97
20
00
20
03
Maddison-Wu (% p.a.)
Off icial (% p.a.)
 
Source for Table 8 Agriculture 1952-90 from Maddison (1998), 1990-2003 from 
official index. Industry 1952-2003, Wu estimates, see table 4. Construction 1952-
2003, official estimates. Transport and communication and commerce 1952-57 from 
Liu and Yeh (1965), thereafter official estimates. Non-material services assumed to 
move parallel to growth in employment (see table 6). 1987 weights were used for 
these Maddison–Wu estimates. 
Source for Table 9 1952-78 from Historical National Accounts of the People’s 
Republic of China (1997) pp. 61 and 70-71. This was a joint study of SSB (State 
Statistical Bureau as NBS was then called) and the Institute of Economic Research of 
Hitotsubashi University in Japan. It provided SNA type estimates for the volume 
movement in each sector and for the value of output in each sector at current prices. 
Here they are updated to 2003, using the NBS Yearbook (2006, p. 60) for volume 
movement in each sector. SSB-Hitotsubashi did not show estimates at constant prices 
as we have in table 9, nor did it explain what weights it used to arrive at its total GDP 
index. Although the volume movement in each of the six sectors in table 9 is the same 
as in SSB-Hitotsubashi, their GDP index showed faster growth than our aggregate 
GDP estimates at 1987 prices. This is due to the fact that the official GDP index was 
constructed in linked segments (see Xu and Ye (2000, p. 18). NBS does not explain 
what weights it uses to arrive at its total GDP index for 1978-2003; however, it shows 
 13 
a volume movement identical with that in Table 9, where 1987 weights are used. Both 
show an average annual growth of 9.59 per cent. 
TABLE 8 MADDISON-WU ESTIMATES OF CHINESE GDP BY SECTOR, 1952-2003 
 GDP in million 1987 yuan GDP in 
 
Agriculture Industry Construction 
Transportation 
& 
Communication 
Commerce “Non material” Services Total GDP 
million 
1990G-K $ 
1952 127,891 17,796 3,658 5,183 14,272 45,486 214,286 305,854 
1953 130,139 22,615 4,990 5,406 14,730 47,038 224,918 321,030 
1954 132,229 26,373 4,821 5,679 15,173 48,014 232,289 331,550 
1955 142,595 27,616 5,487 5,852 15,498 48,803 245,851 350,908 
1956 149,135 33,926 9,238 6,447 16,472 53,042 268,260 382,892 
1957 153,649 38,220 8,662 6,695 16,916 59,877 284,019 405,386 
1958 154,538 58,569 12,993 9,827 17,522 62,512 315,961 450,977 
1959 130,265 79,592 13,728 12,874 18,555 65,264 320,278 457,139 
1960 109,107 87,155 13,919 14,213 16,927 68,136 309,457 441,694 
1961 110,965 47,272 4,821 9,237 12,359 71,135 255,789 365,092 
1962 116,172 40,990 5,970 7,488 11,865 74,266 256,751 366,465 
1963 129,505 47,439 7,514 7,368 12,830 77,535 282,191 402,776 
1964 146,495 56,319 9,434 7,761 14,525 80,961 315,495 450,312 
1965 161,098 69,902 10,433 10,441 14,446 85,227 351,547 501,769 
1966 173,034 83,550 11,413 11,521 17,398 87,610 384,526 548,841 
1967 176,576 69,624 10,846 9,907 18,106 88,654 373,713 533,407 
1968 174,153 66,705 8,794 9,677 16,433 90,469 366,231 522,728 
1969 175,885 87,236 11,826 11,878 19,587 91,218 397,630 567,545 
1970 189,751 115,113 15,422 13,871 21,417 90,673 446,247 636,937 
1971 193,604 129,437 17,295 15,027 21,406 91,694 468,463 668,646 
1972 192,235 139,13916,929 16,471 23,280 94,099 482,153 688,186 
1973 209,868 152,188 17,500 17,500 25,391 95,597 518,044 739,414 
1974 218,892 149,357 18,583 17,555 24,874 97,548 526,809 751,924 
1975 223,928 170,306 21,151 19,562 24,841 99,545 559,333 798,346 
1976 220,352 167,385 22,053 19,246 23,909 103,040 555,985 793,568 
1977 215,841 189,968 22,420 21,679 27,119 113,659 590,686 843,097 
1978 225,079 219,314 22,292 23,617 33,383 131,448 655,133 935,083 
1979 238,994 236,940 22,731 25,432 36,312 145,245 705,654 1,007,193 
1980 235,798 248,837 28,810 26,876 35,841 153,277 729,439 1,041,142 
1981 252,451 253,557 29,722 27,389 46,594 163,330 773,043 1,103,378 
1982 281,773 270,242 30,739 30,589 48,424 169,433 831,200 1,186,387 
1983 305,265 291,058 35,984 33,648 59,020 176,740 901,715 1,287,034 
1984 345,075 323,239 39,891 38,695 71,704 195,369 1,013,973 1,447,262 
1985 351,680 364,042 48,747 43,903 92,392 217,901 1,118,665 1,596,691 
1986 363,504 394,974 56,484 49,519 102,180 226,955 1,193,616 1,703,670 
1987 381,013 460,943 66,580 54,490 115,930 240,320 1,319,276 1,883,027 
1988 390,373 501,043 71,899 61,756 132,475 254,910 1,412,456 2,016,024 
1989 402,216 517,891 65,826 64,669 121,453 265,620 1,437,675 2,051,813 
1990 431,708 528,408 66,609 70,205 115,672 275,400 1,488,002 2,123,852 
1991 441,895 584,966 72,984 78,054 120,880 287,268 1,586,047 2,263,794 
1992 462,722 661,456 88,321 86,249 136,670 304,853 1,740,271 2,483,921 
1993 484,455 752,626 104,215 98,790 148,187 320,442 1,908,715 2,724,344 
1994 503,923 849,759 118,482 110,244 162,308 355,068 2,099,784 2,997,060 
1995 529,052 1,071,138 133,172 125,808 174,827 383,182 2,417,179 3,450,084 
1996 555,991 1,036,432 144,497 142,883 187,345 399,815 2,466,963 3,521,141 
1997 575,460 1,089,751 148,286 161,257 206,774 415,403 2,596,931 3,706,647 
1998 595,608 1,012,770 161,662 178,332 222,798 433,261 2,604,431 3,717,352 
1999 612,360 1,109,987 168,572 202,209 242,995 439,320 2,775,443 3,961,441 
2000 627,075 1,270,707 178,135 229,793 267,498 452,984 3,026,192 4,319,339 
2001 644,732 1,490,760 190,218 256,433 292,635 474,718 3,349,496 4,780,797 
2002 663,522 1,798,983 206,937 281,821 321,645 492,212 3,765,120 5,374,025 
2003 679,821 2,246,790 231,926 305,202 356,931 514,495 4,335,165 6,187,983 
 14 
TABLE 9 OFFICIAL ESTIMATES OF CHINESE GDP BY SECTOR, 1952-2003 
 (million 1987 yuan) GDP in 
 
Agricult
ure 
Industry Construction 
Transportation 
& 
Communication 
Commerce 
Non-
material 
services 
Total 
GDP 
million 
1990 G-K 
$ 
1952 112,038 11,111 3,658 3,637 11,225 13,879 155,548 231,550 
1953 114,167 15,077 4,990 4,513 15,490 16,597 170,834 254,305 
1954 116,072 17,988 4,821 5,004 15,771 15,336 174,992 260,495 
1955 125,259 19,177 5,487 5,128 15,749 16,877 187,677 279,378 
1956 131,085 24,666 9,238 6,244 17,096 19,923 208,252 310,006 
1957 135,118 27,465 8,662 6,695 16,916 21,883 216,739 322,640 
1958 135,679 42,131 12,993 9,827 17,522 26,564 244,716 364,287 
1959 114,167 54,409 13,728 12,874 18,555 30,870 244,603 364,119 
1960 95,457 57,753 13,919 14,213 16,927 34,693 232,962 346,790 
1961 96,913 35,198 4,821 9,237 12,359 27,459 185,987 276,862 
1962 101,283 30,521 5,970 7,488 11,865 25,178 182,305 271,381 
1963 112,711 34,587 7,514 7,368 12,830 26,294 201,304 299,663 
1964 127,276 43,443 9,434 7,761 14,525 31,658 234,097 348,479 
1965 139,600 54,664 10,433 10,441 14,446 37,820 267,404 398,061 
1966 149,683 67,653 11,413 11,521 17,398 31,731 289,399 430,803 
1967 152,484 57,409 10,846 9,907 18,106 32,983 281,735 419,394 
1968 150,132 52,675 8,794 9,677 16,433 35,721 273,432 407,034 
1969 151,364 70,053 11,826 11,878 19,587 38,233 302,941 450,961 
1970 163,016 94,718 15,422 13,871 21,417 39,062 347,506 517,301 
1971 166,041 106,384 17,295 15,027 21,406 42,481 368,634 548,753 
1972 164,585 114,484 16,929 16,471 23,280 42,926 378,675 563,700 
1973 179,374 124,539 17,500 17,500 25,391 44,204 408,508 608,110 
1974 186,768 125,761 18,583 17,555 24,874 46,220 419,761 624,861 
1975 190,577 145,871 21,151 19,562 24,841 48,770 450,772 671,024 
1976 187,216 141,338 22,053 19,246 23,909 50,583 444,345 661,457 
1977 183,071 161,715 22,420 21,679 27,119 53,572 469,576 699,016 
1978 190,577 188,214 22,292 23,617 33,383 58,972 517,055 769,694 
1979 202,341 204,513 22,731 25,432 36,312 62,290 553,619 824,123 
1980 199,316 230,401 28,810 26,876 35,841 70,293 591,537 880,569 
1981 213,209 234,401 29,722 27,389 46,594 74,330 625,645 931,342 
1982 237,858 247,934 30,739 30,589 48,424 87,767 683,311 1,017,184 
1983 247,476 272,033 35,984 33,648 59,020 100,188 748,349 1,114,001 
1984 290,852 312,442 39,891 38,695 71,704 120,073 873,657 1,300,535 
1985 296,118 369,339 48,747 43,903 93,392 137,697 989,196 1,472,528 
1986 305,977 404,949 56,484 49,519 102,180 155,122 1,074,231 1,599,112 
1987 320,430 458,580 66,580 54,490 115,930 180,240 1,196,250 1,780,751 
1988 328,497 528,521 71,899 61,756 132,475 203,045 1,326,193 1,974,186 
1989 338,692 555,242 65,826 64,669 121,453 228,880 1,374,762 2,046,486 
1990 363,453 573,852 66,609 70,205 115,672 236,943 1,426,734 2,123,852 
1991 372,006 656,490 72,984 78,054 120,880 246,481 1,546,895 2,302,725 
1992 389,539 795,392 88,321 86,249 136,670 293,388 1,789,559 2,663,957 
1993 407,835 955,186 104,215 98,790 148,187 330,870 2,045,083 3,044,333 
1994 424,224 1,135,871 118,482 110,244 162,308 368,507 2,319,636 3,453,036 
1995 445,378 1,295,289 133,172 125,808 174,827 402,118 2,576,592 3,835,543 
1996 468,057 1,457,341 144,497 142,883 187,345 437,784 2,837,907 4,224,540 
1997 484,447 1,622,286 148,286 161,257 206,774 480,552 3,103,602 4,620,056 
1998 501,408 1,766,577 161,662 178,332 222,798 517,330 3,348,107 4,984,029 
1999 515,511 1,917,148 168,572 202,209 242,995 557,416 3,603,851 5,364,732 
2000 527,898 2,104,797 178,135 229,793 267,498 601,564 3,909,685 5,820,000 
2001 542,763 2,287,177 190,218 256,433 292,435 661,303 4,230,329 6,297,315 
2002 558,391 2,515,292 206,937 281,822 321,645 732,572 4,616,659 6,872,410 
2003 572,302 2,836,009 231,926 305,202 356,901 801,926 5,104,266 7,598,267 
 
 15 
(VII) Major Conclusions 
1) As the Chinese currency is greatly undervalued, it is particularly important to 
measure its GDP level using a purchasing power (PPP) converter. In 2003, its GDP 
was about three-quarters of the US level. With exchange rate conversion, it was only 
15 percent. In fact, all valid cross-country, regional or world GDP comparisons 
require PPP converters. 
 
2) Chinese official statistics exaggerate the pace of growth by use of dubious deflators. 
Our revisions show GDP growth of 7.85 per cent a year 1978-2003, compared with 
the official rate of 9.59 per cent. In making adjustments for China, we follow normal 
practice in eliminating idiosyncrasies in official statistics to enhance the validity of 
inter-temporal and cross-country comparisons. 
 
3) The two major areas where we amended the official estimates were industry and 
the measurement-resistant part of the service sector. The alternative Wu measure of 
industrial performance is based on quantitative indicators. It is specified in great detail 
and seems clearly superior to the official estimates for this sector, which accounted 
for 55 per cent of official GDP in 2003. We feel that NBS should give serious 
consideration to using the Wu approach to improve its estimate for this sector. 
 
4) Our second major adjustment, for “non-material” services, is much cruder than our 
industry adjustment. This sector represented 15.6 per cent of official GDP in 2003.We 
assumed that there was no productivity growth in this sector and used employment as 
an indicator of the growth of value added. The implicit official estimate for labour 
productivity in this sector was 5.1 per cent a year for 1978-2002, which is higher than 
the rate for other services (transport and commerce combined) where productivity 
grew4.8 per cent a year, and for agriculture, where productivity rose 4 per cent a year. 
The relative pace of growth seems implausible and out of line with the experience of 
other countries (see Table 5). It would be very useful to have more detail on the 
official deflation procedures for this sector. At the moment it is something of a pariah 
in the official accounts; it is not shown explicitly, but as a residual in the official 
estimate for the tertiary sector. 
 
5) We strongly recommend that the official statement of the GDP accounts in the 
China Yearbook be expanded. At the moment information is provided on GDP in 
current prices and volume movement. It would be very useful to have a table showing 
GDP and its six component sectors in constant prices (as in Tables 8 and 9). Such a 
change would require no research, but add considerably to the transparency of the 
accounts. 
 
6) Our estimates suggest quite strongly that official GDP statistics were subjected to a 
smoothing procedure to disguise a significant slowdown of output growth in 1996-
1998 (see Table 2). 
 
7) The official estimates of employment in the China Yearbook have deteriorated 
significantly in the last four years. The 16 sector breakdown which was available for 
1978-2002 has been discontinued, and in its place there is only a rough 3 sector 
breakdown. The large discrepancy between them has been disguised instead of being 
explained. In the present situation, meaningful measurement of labour productivity is 
no longer possible. 
 16 
 
Bibliography: 
DNEA (Department of National Economic Accounts) (1996), Input-Output Table of China 
1992, China Statistical Publishing House. 
DNEB (Department of National Economic Balance, State Statistical Bureau of China) and 
ONIOS (Office of the National Input-Output Survey) (1991), Input-Output Table of 
China 1987, China Statistical Publishing House, Beijing. 
EU, IMF, OECD, UN and World Bank (1993), System of National Accounts, New York, 
Paris etc. 
Liu, T. C. and K. C. Yeh (1965), The Economy of the Chinese Mainland: National Income 
and Economic Development, 1933-59, Princeton University Press. 
Maddison A. (1970), Economic Progress and Policy in Developing Countries, Allen and 
Unwin, London, Norton, New York 
Maddison A. (1995), Monitoring the World Economy, 1820-1992, OECD, Paris. 
Maddison A. (1998a), Chinese Economic Performance in the Long Run, OECD, Paris, out 
of print, available on: www.ggdc.net/Maddison/ 
Maddison A. (1998b), “Measuring the Performance of a Communist Command Economy: An 
Assessment of C.I.A. Estimates for the USSR”, Review of Income and Wealth, 
September, pp. 307-325. 
Maddison A. (2001), The World Economy: A Millennial Perspective, OECD, Paris. 
Maddison A. (2003), The World Economy: Historical Statistics, OECD, Paris. 
NBS, Statistical Yearbook of China, Beijing, various annual issues. 
Ren, R. (1997), China’s Economic Performance in International Perspective, OECD, Paris. 
SSB-Hitotsubash University Institute of Economic Research (1997) The Historical National 
Accounts of the People’s Republic of China, 1952-1995. 
Van Ark B. and N. Crafts (eds) (1996), Quantitative Aspects of post-war European 
Economic Growth, Cambridge University Press. 
Van Ark, B. (1996), “Sectoral Growth Accounting and Structural Change in Postwar Europe” 
pp. 84-164 in van Ark & Crafts 
Wu, H. X. (2000), “Measuring China’s GDP level and growth performance: Alternative 
estimates and the implications”, Review of Income and Wealth, Series 46 (4): 475-499. 
Wu, H. X. (2002), “How Fast has Chinese Industry Grown?-Measuring the Real Output of 
Chinese Industry, 1949-97”, Review of Income and Wealth, June, pp. 179-204. 
Wu, H. X. and X. Yue (2000), “Reconstructing the Post-War Chinese Industrial GDP with a 
Laspeyres’ Quantity Index Approach: A Further Inquiry,” in Konosuke Odaka, 
Yukihiko Kiyokawa and Masaaki Kuboniwa (eds.), Constructing A Historical 
Macroeconomic Database for Trans-Asian Regions, Institute of Economic Research, 
Hitotsubashi University, pp. 85-108 
Xu, X and Y. Ye (2000), National Accounts for China: Sources and Methods, OECD, Paris. 
Yue, X. (2005)”Problems of current Employment Statistics in China”, Economic Research 
Journal (Jingji Yanjiu), March. 
Yue, X., Zhang, S. and X Xu (eds) (2004), Zhongguo Jingji Zengzhang Sudu: Yangjiu Yu 
Zhenglun, Citicpub, Beijing.